Dataset Title: Data from: "Variation of vibratory performance during courtship of a cursorial spider"
===============================================

Corresponding author: Morgan M. Oberweiser
Other authors: Anne G. Hertel, Monika J.B. Eberhard
Date created: 2025-03-11

===============================================
Dataset contents:
-----------------
1. Oberweiser_revised dataset.csv
   - This file contains a selection of vibrational measurement data compiled from the output of the Pulse Train Analysis feature completed with Avisoft Bioacoustics software
   - Details on how this data was revised from the raw PTA output is detailed in the Methods section of the manuscript:

		"These raw measurements were then revised in order to eliminate biologically irrelevant artifacts generated by 
		 the automated pulse interval measurement (i.e. instances where long periods of silence were measured as pulse 
		 intervals). All pulse interval measurements outside of the interquartile range around the median of all pulse 
		 intervals across all trials were therefore excluded."

   - Rows show data on the pulse-level, with individual-level and trial-level information repeated per row for pulses which occured during the same trial or in the same individual
   - Columns contain the following (including examples from the first row of data) :
      - trialID = an ID per trial performed (i.e. "002-1"), comprised of the individual's ID number (i.e. "002"), and the number of the trial out of three (i.e. "1"), separated by a dash
      - ind = the individual spider's ID number (i.e. "2" (simplified in Microsoft Excel from "002"))
      - trial = the number of the trial out of three (i.e. "1")
      - interval = the duration of the pulse measured in seconds (i.e. "0.627") (see Figure 1 in the manuscript for a visual depiction)
         - a PER-PULSE measurement
      - call.dur = the call duration measured in seconds (i.e. "26.174"), calculated as a sum of all pulse intervals occuring during a specific trial (see Figure 1 in the manuscript for a visual depiction)
         - a PER-TRIAL measurement
      - pulse.rt = the pulse rate measured in pulses/second (i.e. "1.681058"), calculated in only the pulses which occured during the calling duration (see Figure 1 in the manuscript for a visual depiction)
         - a PER-TRIAL measurement
      - age = the age, measured in days since final molt, at which the individual spider was tested (i.e. "14")
      - wt = the weight of the individual spider, measured in grams, as measured on the day which the spider was tested (i.e. "0.0638")

2. Oberweiser_statistical code.R
   - This file contains the code for completing the statistical analysis
   - Content is divided into 3 parts:
      - Part 1/3: SUMMARY STATISTICS
           - Individual-based summary stats
           - Trial-based summary stats
           - Pulse-based summary stats
      - Part 2/3: MULTIVARIATE MIXED EFFECTS MODEL ON CALLING DURATION AND PULSE RATE
           - Fit and check model
               - This model takes a small amount of time to run, so code is included to simply load an .rds file, also included in the dataset (3. VibrationMultivariate.rds)
           - Calculate repeatability and correlation
           - Estimate individual behavioral types - calling duration
           - Estimate individual behavioral types - pulse rate
      - Part 3/3: DOUBLE HIERARCHICAL MIXED EFFECTS MODEL ON PULSE INTERVAL
           - Fit and check model
               - This model takes several hours to run, so code is included to simply load an .rds file, also included in the dataset (4. VibrationDHGLM_2025.rds)
           - Calculate repeatability and correlation
           - Estimate individual behavioral types - pulse interval 
           - Estimate individual predictability - pulse interval 

3. VibrationMultivariate.rds
   - This file contains the output of a multivariate Bayesian regression model fitted using the brms package in R
   - This can be loaded in to the above R code (2. Oberweiser_statistical code.R) to avoid processing time for the model in part 2 (MULTIVARIATE MIXED EFFECTS MODEL ON CALLING DURATION AND PULSE RATE)

4. VibrationDHGLM_2025.rds
   - This file contains the output of a Bayesian regression model fitted using the brms package in R
   - This can be loaded in to the above R code (2. Oberweiser_statistical code.R) to avoid processing time for the model in part 3 (DOUBLE HIERARCHICAL MIXED EFFECTS MODEL ON PULSE INTERVAL)